Establishment of an analysis tool for preclinical evaluation of PET radiotracers for in-vivo imaging in neurological diseases

Abstract

Aim: Analysis of in vivo acquired data remains a challenging issue in preclinical studies using positron emission tomography (PET). The aim of this study is the implementation of a tool which should allow a semi-automated analysis of PET data independently from the administrated radiotracer and the imaging modality for preclinical investigations. By registering anatomical data sets, it additionally shall offer a more detailed data analysis allowing a statistical analysis also for smaller sub-regions. Materials and methods: Data used for primarily implementation of the tool were acquired on a Siemens µPET scanner (Inveon®) and were based on the investigation of the glucose metabolism in a subarachnoid hemorrhage (SAH) model in Sprague Dawley rats in vivo using FDG-PET. We used the software programs Matlab (version 2016a) and Fiji for data analysis and visualization. In addition, statistical tests were performed in order to determine regions with trending/significant differences in the SUV of sham and SAH animals. Results: Following data import, data were separated into predefined time periods and artefacts were eliminated. Afterwards, a volume of interest (VOI) was defined by the threshold of the Standardized-Uptake-Value (SUV). Before masking each data set with its segmented VOI, all data sets were intensity-normalized, eliminating the full body intensity differences caused by the different amount of injected activity. After masking, data sets of the sham operated animals were registered on the best orientated sham data set, reduced on the VOI and shifted into the center of the 3D space. By averaging aligned data sets based on all data sets from seven sham rats, we generated a FDG-template of a Sprague-Dawley rat brain. This PET data template was the basis for the evaluation of registered data sets. Afterwards, an anatomical MR-based atlas of the brain of Sprague-Dawley rat was co-registered on the template for a better sub-classification of the acquired data. Conclusion: These preliminary data show that the described method represents a very promising tool for data analysis in the preclinical evaluations of PET radiotracers for neurological applications.